Centroid Community CNC Support Forum

Pete Rondeau is video blogging his Acorn build on his Dyna Myte 2800 bed mill.

thanks Keith!

Pete, this post will help you setup your turns ratio, check backlash etc:
viewtopic.php?f=63&t=1801&p=29005#p29005

Dialing in the Y axis and I go down a backlash rabbit hole.

Pete, Hello! great video! nice job. Glad you found that ball bearing!

back lash is often blamed on the ballscrew but most often is caused by the ballscrew support bearings.
If you can afford it..use “matched set” angular contact bearings for best results.

Here is my backlash measuring method:

1.) slow jog the indicator into the machine to load the indicator, upon contact with the indicator, switch to incremental jogging mode..keep incrementally jogging at .01" per increment until the indicator lands on a dial number that is easy to remember. write down that number...then...
2.) incremental jog away (opposite direction as step one) at .01" per increment at least 8-10 times, count the number of times you hit the incremental jog button.. (note: at this point it does not matter what is being displayed on the indicator so, ignore it..the indicator does not have to be perfectly aligned with axis for this measurement)
3.) now return the indicator to the position made in step one, by pressing the jog button (still in incremental .01" mode) by counting the same number of presses made in step 2.
4.) observe the indicator the amount that the indicator falls short of when trying to return to the indicator reading made in step one. This is the backlash of the axis system.

Discussion:

This backlash measurement method:

Minimizes the axis way sticktion effect from effecting the backlash measurement results.

Secondary benefits: loads the indicator to eliminate play in indicator system and stand before making measurement and does not require the indicator to be aligned perfectly with the axis movement.

Be sure to measure backlash with the backlash compensation value set to zero!

One minor point from the latest video: 25.4 is an EXACT conversion between millimeters and inches. It is not an approximation.

Interesting.. somewhere I got it in my head that it was a long trailing decimal number but I see that you are absolutely correct.. So, you will have to excuse me making the same wrong statement in the just released video! LOL
Thanks for setting me straight.

Finally got to the bottom of the backlash issue. I was lulled into a false sense of security by the number of bearings I originally got out of the ball nut. End up swapping out the thrust bearings, twice- and repack the ball nut with additional bearings. Plus a peak at the newly received Ether1616 expansion board.



Pete

Hmm. The matched bearings are supposed to be fitted in direct contact with each other. That way they should acquire the correct preload without any messing about. You can usually select the degree of preload according to the application by specifying the correct part number. By tightening the pulley as you show, you can't tell how much preload there is.

The bearing housing you are using should ideally be modified so these nice NSK can be used as intended, by removing the step in the bore and fitting an external feature for the bearings to sit up against and a retainer plate on the other side to squash them together

Not sure why adding an extra 4 balls would make any measurable difference to the backlash. Fitting larger diameter balls would be a different matter.

Yes, these bearings are basically set up the way you would install pinion bearings into an automotive differential. With as much as I have on my plate, I'm not really keen on the idea of redesigning everything.

I agree with you on the balls.. In my mind the actual ball count shouldn't even matter but in reality, there is a huge difference that you can both measure and feel when turning the screw.

Pete